The capture ofDiaptomus siciloidesandDiaptomus pallidusby gizzard shad (Dorosoma cepedianum) was simulated using a suction device. Based on simulations, the capture probability forD.siciloideswas significantly greater (p < 0.025) than that forD.pallidus. Capture probabilities in the simulations varied from 0.092–0.391 forD.siciloidesand from 0.047–0.210 forD.pallidus. These capture probabilities usually underestimated the capture frequencies observed in feeding experiments employing live shad. The shad did not exhibit a tendency to selectD.siciloidesoverD.pallidusbut appeared to capture the calanoids in proportion to their densities in the test pools.